Sadness

Abstract

Sadness, which MeSH classifies under emotions, is the negative emotion that typically follows distressing and adverse situations such as the loss of someone or the failure to achieve a goal. This article covers how sadness is defined within basic emotion theory, how it is elicited in the laboratory, its facial expression and autonomic signature, its brain correlates, its effects on judgment, decision making and spending, why sad music can be pleasurable, and how transient sadness relates to depression. The central dispute is whether sadness is a discrete emotion with its own neural and physiological signature or a category constructed from more general brain networks, with a recent review proposing that the two views may be complementary.

Keywords: sadness, basic emotions, negative affect, emotion and decision making

Sadness is generally seen as a negative emotion, a response to distressing and adverse situations (Sachs et al., 2015). It is also one of the emotions on which competing theories of emotion have been tested, since it can be elicited in the laboratory and its transient form overlaps with the mood disturbance of depression.

What Sadness Is

Basic emotion theory treats sadness as one member of a small evolved set. In that account, emotions evolved through their adaptive value in dealing with fundamental life-tasks, each emotion has unique features of signal, physiology and antecedent events, and all share rapid onset, short duration, unbidden occurrence, automatic appraisal and coherence among responses (Ekman, 1992). The antecedent events that distinguish sadness from neighbouring states have been studied directly. When 445 participants identified depressive symptoms that followed a recent adverse situation, crying, sadness and desire for social support were prominent following social losses, whereas guilt, rumination, fatigue and pessimism were prominent following failed efforts (Keller & Nesse, 2006).

Whether sadness is a single emotion is itself an open question. In an experiment that compared sadness induced by the loss of someone with sadness induced by failure to achieve a goal, loss-evoked sadness was associated with expressive words relating to tears, skin conductance took longer to return to baseline after loss, and the authors suggested that sadness includes at least two subtypes (Shirai & Suzuki, 2017).

Eliciting and Measuring Sadness

Laboratory research on sadness depends on mood induction. One programme evaluated over 250 films and selected the two best films for each of eight target emotions, and the resulting set of 16 films successfully elicited sadness among other states in an ethnically diverse sample of 494 subjects (Gross & Levenson, 1995). A meta-analysis of 687 studies with 49,473 participants that elicited happiness, sadness, anger and anxiety found that most emotion elicitations were effective, that picture presentations were overall the most effective elicitor, and that a greater proportion of women was associated with stronger effects for elicitations of happiness and sadness (Lench et al., 2011).

The same meta-analysis bears on what sadness is. Consistent with discrete emotion theory, it found moderate differences among discrete emotions, differences among discrete negative emotions, and correlated changes in behaviour, experience and physiology, while valence, valence-arousal and approach-avoidance models were not as clearly supported (Lench et al., 2011). Self-report data point the other way on boundaries: analysis of reports elicited by 2,185 short videos uncovered 27 varieties of reported emotional experience, many of them linked by smooth gradients, contrary to discrete theories (Cowen & Keltner, 2017).

Expression and Physiology

Sadness has a characteristic bodily form. It is typically characterized by raised inner eyebrows, lowered corners of the mouth, reduced walking speed and slumped posture (Arias et al., 2020). Cross-cultural work tested whether such facial patterns are learned from media: members of a preliterate culture in New Guinea, told a story and shown three faces, selected the face appropriate to the story, providing evidence that the association between particular facial muscular patterns and discrete emotions is universal (Ekman & Friesen, 1971). Later modelling of dynamic expressions questioned the number of categories involved, finding that facial expressions transmit an evolving hierarchy of signals over time and suggesting four rather than six basic categories (Jack et al., 2014). The universality claim has also been challenged on method: one review argued that forced-choice response formats, within-subject designs and preselected photographs of posed expressions are each problematic, and that when they are altered, less supportive or nonsupportive results occur (Russell, 1994). A later review concluded that people do sometimes frown when sad more than would be expected by chance, but that how people communicate sadness varies substantially across cultures, situations and even across people within a single situation (Barrett et al., 2019).

Peripheral physiology is less settled. A review of 134 publications on emotional effects on peripheral physiological responding in healthy individuals suggested considerable autonomic response specificity in emotion when considering subtypes of distinct emotions (Kreibig, 2010). In the loss and failure subtypes of sadness, blood pressure changes might be influenced by different factors: tear ratings correlated with blood pressure in the loss condition, whereas sadness intensity correlated with blood pressure in the failure condition (Shirai & Suzuki, 2017).

The Brain and Sadness

The best-known imaging result on sadness links it to depression. Using positron emission tomography, provocation of transient sadness in healthy volunteers increased limbic-paralimbic blood flow in the subgenual cingulate and anterior insula and decreased activity in right dorsolateral prefrontal and inferior parietal regions, whereas recovery from depression showed the reverse pattern in the same regions (Mayberg et al., 1999). A later synthesis described ancient subcortical circuitry extending from dorsal periaqueductal grey to subgenual anterior cingulate, the latter now a treatment target in disorders of sadness, and noted a role for reduced left relative to right frontal asymmetry in sadness (Arias et al., 2020).

Whether such findings show a dedicated sadness circuit is disputed. A voxel-based meta-analysis found that each of the emotions examined, including sadness, was characterized by consistent neural correlates across studies, and that the activation patterns were discriminable from the other emotions in pairwise contrasts (Vytal & Hamann, 2010). A second meta-analysis found little evidence that discrete emotion categories can be consistently and specifically localized to distinct brain regions, and evidence instead for a set of interacting regions involved in basic psychological operations of both an emotional and non-emotional nature (Lindquist et al., 2012). The latter view was developed into a brain-based, computational account called the theory of constructed emotion (Barrett, 2017). A collaborative review of the sadness literature suggested that the field may be moving toward a theoretical consensus in which basic emotion theory and psychological constructionism are complementary, working at different levels of the phylogenetic hierarchy (Arias et al., 2020).

Sadness and Cognition

Negative affect has measurable cognitive benefits. A review of experimental evidence reported that negative affect can improve memory performance, reduce judgmental errors, improve motivation and result in more effective interpersonal strategies, and interpreted these effects in terms of dual-process theories in which negative affect promotes a more accommodative and externally focused thinking strategy (Forgas, 2013). A later review from the same laboratory interpreted its results in terms of a dual-process theory predicting that negative affect promotes a more accommodative, vigilant and externally focused thinking strategy, and described the benefits of negative affect as consistent with evolutionary theories of the adaptive signalling function of affective states (Forgas, 2017). An earlier integrative account, the affect infusion model, predicts that the degree to which mood infuses judgments varies along a processing continuum, so that judgments requiring heuristic or substantive processing are more likely to be infused by affect than direct-access or motivated judgments (Forgas, 1995).

Sadness also differs from other negative emotions in what it does to judgment. In two experiments, sad participants perceived situationally caused events as more likely and situational forces as more responsible for an ambiguous event than angry participants, who perceived events caused by humans as more likely and other people as more responsible (Keltner et al., 1993). In gambling and job-selection decisions, sad individuals were biased in favour of high-risk, high-reward options, whereas anxious individuals favoured low-risk, low-reward options; the authors argued that sadness primes an implicit goal of reward replacement (Raghunathan & Pham, 1999).

Spending shows the same specificity. The endowment effect is the tendency for selling prices to exceed buying or choice prices for the same object; as predicted by appraisal-tendency theory, disgust induced by a prior, irrelevant situation carried over to such decisions and eliminated it, whereas sadness also carried over but reduced selling prices and increased choice prices, producing a reverse endowment effect in which choice prices exceeded selling prices (Lerner et al., 2004). A follow-up found that sadness increases the amount of money decision makers give up to acquire a commodity, but only when self-focus is high, so that self-focus moderates the effect of sadness on spending (Cryder et al., 2008).

The Pleasure of Sad Art

Sadness is often associated with some degree of pleasure in an aesthetic context, as the popularity of sad music, plays, films and paintings suggests (Sachs et al., 2015). A systematic review proposed that sadness evoked by music is found pleasurable when it is perceived as non-threatening, when it is aesthetically pleasing, and when it produces psychological benefits such as mood regulation and empathic feelings (Sachs et al., 2015). A trait account adds that not all listeners enjoy sad music: enjoyment is associated with high empathic concern and imaginative absorption with only nominal personal distress, although the correlations are modest (Huron & Vuoskoski, 2020).

Such mixed responses bear on the structure of emotion. Although most participants surveyed in typical situations felt either happy or sad, many surveyed immediately after watching a tragicomic film, moving out of their dormitories or graduating from college felt both happy and sad, a pattern the authors took to favour the evaluative space model, in which positive and negative affect are separable, over a circumplex in which happiness and sadness are polar opposites, concluding that the underlying processes are better characterized as bivariate (Larsen et al., 2001).

Sadness and Depression

One classic study linked depression to more accurate judgment. In four experiments, depressed students' judgments of contingency were surprisingly accurate, whereas nondepressed students overestimated the degree of contingency between their responses and outcomes when noncontingent outcomes were frequent or desired (Alloy & Abramson, 1979). Normal sadness and clinical depression share neural territory. Reciprocal changes involving the subgenual cingulate and right prefrontal cortex occurred with both transient and chronic changes in negative mood (Mayberg et al., 1999). An evolutionary account holds that, although much depression may be dysfunctional, the capacity to experience normal depressive symptoms in response to certain adverse situations appears to have been shaped by natural selection, so that different situations evoke different symptom patterns, which the authors called the situation-symptom congruence hypothesis (Keller & Nesse, 2006). The sadness synthesis suggested that reduced cardiac control may in part contribute to epidemiological reports of reduced life expectancy in affective disorders (Arias et al., 2020).

Worked Example

The spending demonstration above uses illustrative prices, not data, to show the reversal reported by Lerner et al. (2004). With a neutral mood the demonstration sets a selling price of 8 and a choice price of 5, a gap of 8 - 5 = 3 in favour of selling, which is the ordinary endowment effect. With sadness the demonstration lowers the selling price to 6 and raises the choice price to 7, so the gap becomes 6 - 7 = -1: choice prices now exceed selling prices, the reverse endowment effect the study reported (Lerner et al., 2004).

The elicitation meta-analysis can be put on a per-study scale. Its 49,473 participants across 687 studies (Lench et al., 2011) average about 72 participants per study, and its 4,946 effects average about 7 effects per study. The film set selected 16 of the more than 250 films evaluated (Gross & Levenson, 1995), two per target emotion, which is under 7% of the films screened.

Discussion

Three conclusions follow. First, the discrete and constructionist programmes disagree over sadness, as over the other emotions both meta-analyses examined, because the neuroimaging literature yields consistent, discriminable correlates in one meta-analysis (Vytal & Hamann, 2010) and little evidence for consistent and specific localization in another (Lindquist et al., 2012). The suggestion from one experiment that sadness includes at least two subtypes with different responses (Shirai & Suzuki, 2017) cuts across that dispute, since a category with internal subtypes is awkward for a one-emotion, one-signature view.

Second, the decision-making findings show that valence alone is not a sufficient description of an emotion. Sadness and anxiety, both negative, pushed risky choices in opposite directions (Raghunathan & Pham, 1999), and sadness and disgust had opposing effects on choice prices (Lerner et al., 2004); the elicitation meta-analysis likewise found that valence models were not as clearly supported as discrete emotion theory (Lench et al., 2011).

Third, the functional reading of sadness has empirical support in normal-range moods. The experimental evidence that negative affect can improve memory and reduce judgmental errors (Forgas, 2013) concerns negative mood in experiments, and does not by itself bear on the dysfunction of depression, which the evolutionary account explicitly allows (Keller & Nesse, 2006).

Current Directions

Current work runs along three lines. One integrative review of the neuroscience of sadness has suggested that basic emotion theory and psychological constructionism may be complementary, operating at different levels of the phylogenetic hierarchy, and has noted that the subgenual anterior cingulate is now a treatment target in disorders of sadness and that stimulation of left dorsolateral prefrontal cortex is of interest as an antidepressant target (Arias et al., 2020). Large-scale self-report modelling has uncovered 27 varieties of reported emotional experience, many of them linked by smooth gradients (Cowen & Keltner, 2017). Research on aesthetic sadness has also asked who enjoys sad music, with a trait-empathy account suggested to apply to many other situations in which spectators take pleasure in tragic portrayals (Huron & Vuoskoski, 2020).

Common Misconceptions

Sadness is simply a mild form of depression.

Transient sadness in healthy volunteers and recovery from depression involved the same regions changing in opposite directions (Mayberg et al., 1999), and an evolutionary account treats normal depressive symptoms after adverse situations as shaped by selection while allowing that much depression may be dysfunctional (Keller & Nesse, 2006). The overlap is real, but on the evolutionary account normal depressive symptoms and dysfunctional depression are not the same state.

Sadness only impairs thinking.

Experiments have shown that negative affect can improve memory performance, reduce judgmental errors and improve motivation (Forgas, 2013).

All negative emotions affect decisions the same way.

Sad individuals favoured high-risk, high-reward options whereas anxious individuals favoured low-risk, low-reward options (Raghunathan & Pham, 1999).

Glossary

Appraisal-tendency theory.

The account under which specific emotions carry over to unrelated decisions in emotion-specific ways; it predicted the carryover of disgust to pricing decisions.

Basic emotion theory.

The view that a limited set of emotions evolved for fundamental life-tasks, each with its own signal, physiology and antecedent events.

Carryover effect.

The influence of an emotion induced by an irrelevant prior situation on a later, normatively unrelated decision.

Choice price.

The amount of money a person treats as equivalent to receiving an object, used with the selling price to measure the endowment effect.

Endowment effect.

The tendency for selling prices to exceed buying or choice prices for the same object.

Evaluative space model.

The proposal that positive and negative affect are separable, so that mixed feelings of happiness and sadness can co-occur.

Facial expression.

A pattern of facial muscle movement associated with an emotion; for sadness, typically raised inner eyebrows and lowered corners of the mouth, although how people communicate sadness varies substantially across cultures and situations.

Frontal asymmetry.

A difference in electrical activity between left and right frontal cortex; reduced left relative to right activity has been linked to sadness.

Mood induction.

A laboratory procedure, such as a film clip or picture presentation, used to elicit a target emotion.

Psychological constructionism.

The view that emotion categories are constructed from more general brain networks that are not specific to those categories.

Reverse endowment effect.

The pattern, observed under sadness, in which choice prices exceed selling prices.

Reward replacement.

The implicit goal that sadness is proposed to prime, favouring options that offer high rewards.

Situation-symptom congruence hypothesis.

The proposal that different adverse situations evoke different patterns of depressive symptoms suited to the challenge each poses.

Skin conductance level.

A measure of sweat-gland activity used as an index of autonomic arousal.

Subgenual cingulate.

A region of the anterior cingulate cortex whose blood flow rose with induced sadness and whose metabolism fell with recovery from depression.

Theory of constructed emotion.

A brain-based, computational account of what emotions are and how they work, developed from the constructionist view.

Key Researchers

Lisa Feldman Barrett

. Psychologist at Northeastern University; co-author of the meta-analysis that found little evidence for the consistent and specific localization of discrete emotions, and author of the theory of constructed emotion. ORCID - Wikidata - Wikipedia - Google Scholar - Faculty Page

Paul Ekman

(1934-2025). Psychologist at the University of California, San Francisco; argued that each basic emotion has unique features of signal, physiology and antecedent events, and tested the universality of facial expressions in New Guinea. Wikidata - Wikipedia - Google Scholar

Joseph P. Forgas

. Psychologist at the University of New South Wales; reviewed experimental evidence that negative affect can improve memory performance, reduce judgmental errors and improve interpersonal strategies. ORCID - Google Scholar - Faculty Page

James J. Gross

. Psychologist at Stanford University; with Robert Levenson developed a set of films that elicit sadness and other emotions in the laboratory, and co-authored the study linking sadness and self-focus to spending. ORCID - Wikidata - Wikipedia - Google Scholar - Faculty Page

Jennifer S. Lerner

. Researcher at Harvard University; showed that incidental sadness and disgust carry over to economic decisions with opposing effects on choice prices. ORCID

Helen S. Mayberg

. Neurologist at the Icahn School of Medicine at Mount Sinai; reported reciprocal subgenual cingulate and right prefrontal changes shared by induced sadness and recovery from depression. ORCID - Wikidata - Wikipedia - Faculty Page

Frequently Asked Questions

What is sadness in psychology?

Sadness is generally seen as a negative emotion, a response to distressing and adverse situations (Sachs et al., 2015).

What does sadness look like on the face and body?

Sadness is typically characterized by raised inner eyebrows, lowered corners of the mouth, reduced walking speed and slumped posture (Arias et al., 2020), although how people communicate sadness varies substantially across cultures, situations and people (Barrett et al., 2019).

What usually causes sadness?

In a study of 445 participants who identified depressive symptoms after a recent adverse situation, crying, sadness and desire for social support were prominent following social losses (Keller & Nesse, 2006).

Is sadness one emotion or several?

One experiment found that sadness induced by the loss of someone and by failure to achieve a goal produced different responses, suggesting at least two subtypes (Shirai & Suzuki, 2017).

Does the brain have a sadness centre?

Meta-analyses disagree: one found consistent and discriminable neural correlates for sadness and the other emotions it examined (Vytal & Hamann, 2010), while another found little evidence that discrete emotion categories can be consistently and specifically localized to distinct brain regions (Lindquist et al., 2012).

How does sadness affect spending?

Sadness increased the amount of money decision makers gave up to acquire a commodity, but only when self-focus was high (Cryder et al., 2008).

Why do people enjoy sad music?

Enjoyment of sad music is modestly associated with high empathic concern and imaginative absorption with only nominal personal distress (Huron & Vuoskoski, 2020).

Can sadness be useful?

Experiments have shown that negative affect can improve memory performance and reduce judgmental errors (Forgas, 2013).

References